AI Analytics

Technical writing

CFPB Enforcement Actions: The Public Record of $20 Billion in Consumer Finance Penalties

· AI Analytics
Federal DataCFPBConsumer FinanceEnforcement

Since opening its doors in 2011, the Consumer Financial Protection Bureau has brought more than 200 formal enforcement actions against banks, lenders, servicers, and debt collectors — extracting over $20 billion in consumer relief and civil money penalties. The public enforcement database at consumerfinance.gov is the most granular window into how federal consumer financial law is actually applied, case by case, company by company.

Enforcement actions vs. consumer complaints

The CFPB maintains two distinct public datasets that are frequently confused. The Consumer Complaint Database (5M+ complaints) is a real-time firehose of individual grievances submitted by the public — unverified, voluminous, and useful as a leading indicator of industry problems. The Enforcement Actions database is the opposite: a curated, authoritative record of formal legal proceedings that the Bureau has actually filed against specific companies. Where a complaint is an allegation, an enforcement action is a legal finding (or at minimum, a credible public accusation backed by the Bureau's investigative resources).

The pipeline between the two runs in one direction. Complaints accumulate, examiners identify patterns during supervisory exams, and enforcement referrals follow months or years later. Researchers who track complaint volume against specific institutions often see elevated complaint counts 18–36 months before a public enforcement action appears. The enforcement record is therefore the lagging confirmation of what the complaint signal was already saying.

Statutory authority

The CFPB's enforcement powers derive primarily from Title X of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010. Two sections define the litigation framework:

Beyond Dodd-Frank, the CFPB has authority to enforce 19 enumerated federal consumer financial laws, including the Truth in Lending Act (TILA), Real Estate Settlement Procedures Act (RESPA), Equal Credit Opportunity Act (ECOA), Fair Credit Reporting Act (FCRA), Fair Debt Collection Practices Act (FDCPA), Electronic Fund Transfer Act (EFTA), and the Home Mortgage Disclosure Act (HMDA), among others. This jurisdiction covers the full consumer financial lifecycle: origination, servicing, reporting, collections, and remittances.

The UDAAP standard

The most powerful — and most contested — enforcement tool in the CFPB's arsenal is the prohibition against “unfair, deceptive, or abusive acts or practices” (UDAAP). The “unfair” and “deceptive” prongs trace back to the FTC Act and decades of case law. Dodd-Frank added “abusive” as a novel third prong that has no direct predecessor in federal law.

The distinctions matter operationally:

The “abusive” standard gave the CFPB a tool calibrated for modern financial products, where complexity and information asymmetry are often the injury mechanism rather than outright fraud. It has also been the most legally contested prong: the Fifth Circuit's 2024 decision in CFPB v. CURO and subsequent challenges attempted to narrow its scope, with mixed results at the circuit level as of mid-2026.

What the enforcement database contains

Each record in the enforcement actions database includes:

The database is publicly searchable at consumerfinance.gov/enforcement/actions/ but is not available as a bulk CSV download. Each action is a structured HTML page with linked PDFs, which requires scraping or the CFPB's public API to aggregate at scale.

Scale and aggregate statistics

As of mid-2026, the CFPB has brought more than 200 public enforcement actions since its first in 2012. The aggregate relief across all resolved actions exceeds $20 billion, spanning:

The distribution is heavily skewed. The top ten actions by dollar value account for roughly half the total relief. Median action size is substantially lower — many cases against smaller mortgage companies, debt collectors, and credit repair firms resolve in the low millions.

Landmark cases

A handful of enforcement actions illustrate the full range of the CFPB's authority:

Wells Fargo — $3.7 billion (2022)

The largest enforcement action in CFPB history. The December 2022 consent order covered failures across three product lines simultaneously: auto loan servicing (improper repossessions, insurance add-ons, payment processing errors), mortgage origination and servicing (wrongful denials, fee errors, foreclosure processing failures), and deposit accounts (surprise overdraft fees, unlawful account freezes). The $3.7 billion included $2.0 billion in consumer redress and $1.7 billion in civil money penalties. The action followed years of supervisory findings and smaller enforcement actions going back to the 2016 fake accounts scandal.

Navient — $195 million (2022)

The CFPB's January 2022 settlement with Navient, the largest student loan servicer in the country, resolved a lawsuit originally filed in 2017. The Bureau alleged Navient steered borrowers who qualified for income-driven repayment plans into costly forbearance instead, accumulated interest, and misreported loan statuses to credit bureaus. The $195 million settlement included $95 million in restitution to approximately 350,000 federal student loan borrowers and cancellation of $1.7 billion in private loan balances for approximately 66,000 borrowers.

BancorpSouth — $10.6 million (2016)

One of the CFPB's landmark fair lending actions. Examiners found that BancorpSouth denied mortgage applications from Black applicants at rates inconsistent with creditworthiness differences, and that loan officers made racially disparate statements captured in testing. The redlining settlement, coordinated with the DOJ, included $4 million in community investment and required enhanced fair lending compliance infrastructure across the bank's Mississippi, Alabama, and Tennessee markets.

Credit Acceptance Corporation — $27.2 million (2021)

The CFPB sued Credit Acceptance, a major subprime auto lender, alleging it structured loans it knew borrowers could not repay, hid the true cost of financing through add-on products, and misled dealers who originated the loans. The $27.2 million settlement included consumer redress and a civil money penalty, along with requirements to change underwriting and disclosure practices.

MoneyGram — $18 million (2023)

The CFPB's action against MoneyGram targeted remittance transfer failures: the company disclosed incorrect exchange rates and transfer fees, failed to deliver funds within the legally required window, and did not provide required receipts and error resolution disclosures under the Electronic Fund Transfer Act's remittance rules. The $18 million penalty (paid to Treasury) was paired with a compliance order requiring systematic disclosure reforms.

Supervision vs. enforcement: the iceberg dynamic

Enforcement actions are the visible tip of the CFPB's supervisory iceberg. The Bureau has examination authority over all federally insured depository institutions with more than $10 billion in assets (roughly 100 institutions) and, uniquely among federal regulators, over certain categories of nonbank financial companies regardless of size — mortgage companies, payday lenders, private student loan servicers, auto lenders, debt collectors, and prepaid card providers.

Most supervisory findings never become public enforcement actions. Examiners issue confidential supervisory findings and matters requiring attention (MRAs), and institutions typically remediate violations quietly. Public enforcement follows when: (a) the violation is systemic rather than isolated; (b) prior supervisory directives were not followed; (c) the consumer harm is large enough to justify the resource cost of litigation; or (d) public deterrence is a Bureau priority. This means the enforcement database, large as it is, represents a carefully curated selection from a much larger universe of identified violations.

Political cycle and enforcement trends

The enforcement record maps closely to political administrations, which control Bureau leadership through the director appointment:

Accessing the database

The CFPB publishes enforcement actions at consumerfinance.gov/enforcement/actions/. The interface supports filtering by date range, law violated, and company name, but does not offer a bulk CSV download. For analytical work, researchers have two options:

Python: scraping and analyzing enforcement actions

The script below scrapes the CFPB enforcement actions listing, extracts key fields from each action's detail page, parses dollar amounts from prose relief descriptions, and outputs a sorted table plus a CSV export. It uses only the standard library plus requests and beautifulsoup4.

import requests
from bs4 import BeautifulSoup
import re
import csv
import time

BASE_URL = "https://www.consumerfinance.gov"
ACTIONS_URL = BASE_URL + "/enforcement/actions/"

def parse_relief_amount(text):
    """Extract dollar amount from relief text, return float in millions or 0."""
    if not text:
        return 0.0
    # Match patterns like $3.7 billion, $195 million, $27.2 million
    match = re.search(r'\$(\d+(?:\.\d+)?)\s*(billion|million|thousand)?', text, re.IGNORECASE)
    if not match:
        return 0.0
    amount = float(match.group(1))
    unit = (match.group(2) or "").lower()
    if unit == "billion":
        amount *= 1000
    elif unit == "thousand":
        amount /= 1000
    return amount  # in millions

def fetch_enforcement_actions(max_pages=10):
    """Scrape CFPB enforcement actions listing pages."""
    actions = []
    page = 1

    while page <= max_pages:
        params = {"page": page}
        resp = requests.get(ACTIONS_URL, params=params, timeout=30,
                            headers={"User-Agent": "research-bot/1.0"})
        if resp.status_code != 200:
            break

        soup = BeautifulSoup(resp.text, "html.parser")

        # Each action is in an <article> or list item with class activity-block
        items = soup.select("article.o-post-preview, li.o-post-preview")
        if not items:
            break

        for item in items:
            title_el = item.select_one(".o-post-preview__title a, h3 a, h2 a")
            date_el  = item.select_one("time, .o-post-preview__time")
            tags_el  = item.select_all(".o-post-preview__topics li, .m-tag")

            name = title_el.get_text(strip=True) if title_el else ""
            href = title_el["href"] if title_el and title_el.has_attr("href") else ""
            date = date_el.get("datetime", date_el.get_text(strip=True)) if date_el else ""
            tags = [t.get_text(strip=True) for t in tags_el] if tags_el else []

            # Fetch individual action page for relief amount
            relief_text = ""
            if href:
                url = href if href.startswith("http") else BASE_URL + href
                try:
                    detail = requests.get(url, timeout=20,
                                          headers={"User-Agent": "research-bot/1.0"})
                    dsoup = BeautifulSoup(detail.text, "html.parser")
                    # Look for redress / penalty summary paragraph
                    for p in dsoup.select("p, li"):
                        if re.search(r'\$(\d)', p.get_text()):
                            relief_text = p.get_text(strip=True)[:200]
                            break
                except Exception:
                    pass
                time.sleep(0.5)

            actions.append({
                "name": name,
                "date": date,
                "url": href,
                "tags": "; ".join(tags),
                "relief_text": relief_text,
                "relief_millions": parse_relief_amount(relief_text),
            })

        page += 1
        time.sleep(1)

    return actions

def main():
    print("Fetching CFPB enforcement actions...")
    actions = fetch_enforcement_actions(max_pages=5)

    # Sort by relief amount descending
    actions.sort(key=lambda x: x["relief_millions"], reverse=True)

    # Print top 20 as a table
    print("\n{:<50} {:>12} {}".format("Company / Action", "Relief ($M)", "Date"))
    print("-" * 80)
    for a in actions[:20]:
        label = a["name"][:48]
        relief = "{:.1f}".format(a["relief_millions"]) if a["relief_millions"] else "n/a"
        print("{:<50} {:>12} {}".format(label, relief, a["date"][:10]))

    # Export full dataset to CSV
    if actions:
        with open("cfpb_enforcement_actions.csv", "w", newline="") as f:
            writer = csv.DictWriter(f, fieldnames=actions[0].keys())
            writer.writeheader()
            writer.writerows(actions)
        print("\nExported " + str(len(actions)) + " actions to cfpb_enforcement_actions.csv")

if __name__ == "__main__":
    main()

A few implementation notes. The relief amount parser uses a regex that handles both “$3.7 billion” and “$195 million” patterns and normalizes to millions for sorting. The detail-page fetch includes a 500ms delay to avoid overloading the CFPB servers; the outer page loop sleeps one second between pages. For a complete dataset, set max_pages high enough to exhaust the listing (currently ~15 pages at 10 actions per page). The CSV export includes the raw relief text so you can audit the parser's extraction.

Cross-agency context

The CFPB does not operate in isolation. Several other federal agencies run parallel or overlapping enforcement programs that analysts should track alongside the CFPB record:

Analytical applications

The enforcement database is a primary source for several categories of research and due diligence work:

Data quality and limitations

The enforcement database is authoritative but incomplete in a specific sense: it records only public actions. Confidential supervisory findings, informal enforcement guidance, and matters resolved through supervisory channels without public announcement are not included. The database also does not include actions against individuals (as opposed to companies) in most cases, and the relief amounts reported in the database are as-agreed in the consent order, not as-actually-collected — a distinction that matters when an institution later fails or is acquired.

Relief descriptions are prose rather than structured fields in the underlying HTML, which is why the parser in the code above uses regex extraction rather than a clean JSON attribute. Dollar amounts sometimes appear in multiple paragraphs (total relief vs. consumer redress vs. civil money penalty broken out separately), and the parser's first-match logic will capture whichever appears first in the document. For high-precision analysis, manual review of the PDF consent orders is the authoritative source.

Related writing

The CFPB consumer complaint database tracks 5M+ complaints that often precede enforcement action by 18–36 months — the leading indicator to this dataset's lagging record.

The SEC enforcement actions database covers 700–800 securities law actions per year using a similar administrative proceeding / court action structure.

The HHS OIG exclusion list is the healthcare equivalent — a public blacklist of individuals and entities barred from federal program participation.